Relative deflection in wood roof trusses 2

[RontheRedneck]

Had an interesting situation pop up. Thought it might be worth mentioning here in case any of you ran across a similar situation.

The building is a large one. Has some rather shallow 60' plus span scissor trusses on it. The building has a hip roof, which further complicates things.

Either the architect or EOR figured some bearing would be needed on the shallowest trusses, so they put a bearing wall in near the hip end.

The last truss that hits the interior bearing has about 1/4" of deflection. The next truss that's clear span has 3" of deflection. (All deflection numbers are for total load)

The specs call for a total load deflection limit of L/240. On a 60' truss that's roughly 3".

So technically, the trusses work. They meet the deflection criteria in the specs. But no way in hell is that a good idea.


I was not involved in the job until yesterday. A guy asked me to look at the job and see what I thought. I suggested referring it back to the architect and EOR to see what they want to do. That's not a problem we can solve.

 

[DoubleStud]

I wouldn't worry about it. Most of the deflection is caused because of the legs are trying to spread. It will all work together. The truss that has small deflection will keep the walls from spreading more. You will will probably have other interior walls as well that will help.

 

[XR250]

I disagree with DS and agree with RTR.

Regardless of L/XXX, 3" is too much. The EOR needs to figure their shit out. Technically, it is waay more than L/240 as you have about 3" deflection over 24" - not that this is ever enforced.

 

[JAE]

It's not about "not worrying" about it. The huge step in deflection will make the roof line look odd, create issues with ceiling sheathing (if used) and with long term creep in the trusses, grow the deflection difference even more over time. Probably not a safety issue but if it was my building as the owner I'd not want to pay for that.

 

[DoubleStud]

What is the horizontal deflection on the typical/middle trusses?

Technically speaking an end wall or a truss that is embedded in the wall has 0 deflection. So there is always a big deflection difference from the end and first truss especially for scissor trusses. If you model the truss on the software with mid support, then the vertical deflection is probably taken on the 1/4 point, not the mid point. Mid point is probably 0 because it is a support. But then again I have never dealt with 60 ft trusses, you may be right to worry about it. How do you ship this?

 

[hokie66]

Agree with RR and JAE. May not be a strength problem, but certainly a serviceability issue and a severe one. The architect and EOR should know better, and now have a problem.

 

[Eng16080]

That's an unfortunate problem which doesn't appear to have an easy solution, unless they plan on having a drop ceiling. I'll give the engineer a little sympathy on this one though. I can see somebody easily overlooking/missing this.

I'll give one possible solution (half joking): remove the support wall and reconstruct it 3" shorter. The trusses supported on the wall should eventually sag until they land on the wall. Then all the trusses should have the same deflection.

 

[RontheRedneck]

Just to be clear - My intent was not to blame anyone. It's to let others known about the issue and maybe head off problems down the road.

As Eng16080 said, this was easy to overlook. The architect and EOR did not think about it. Apparently the truss designer who quoted the job a couple of months back did not think about it.

The problem was headed off before anything got built. I don't think they even have concrete in the ground yet. That's the best possible outcome given the situation.

 

[Mark_J]

Just tell the EOR to put a beam in the design and be done with it. Many things are overlooked, differential deflection is a common one.

 

[jerseyshore]

Agreed Mark, this is one of the more common "mistakes" that doesn't get picked up until the ceiling starts to crack or residents complain that their unit feels different halfway thru. We worked with a truss designer that marked on the plan for us and ownership to review potential differential deflection issues. Was helpful to have that documented in design phase so that ownership could make a call if they wanted to add more beams/ bearing walls or accept that a difference was expected.

 

[phamENG]

I've read a few case studies after lawsuits over this issue. Usually it's a floor truss that's 8' long next to one that's 18' long, but same basic idea. I know AISC and SJI have covered this issue in webinars in the past. I don't believe IBC deflection criteria have a stated restriction on what they apply to. The plywood sheathing is a roof 'member' and should be held to at least the same standard.

 

[DoubleStud]

Ok, please help me understand, let’s say the project is a gable roof instead of a hip roof, are you all OK with 3” deflection on the common trusses? I would think if you ask them to make them stiffer, they would come back with a change order.

My second question is this, when I have projects with scissor trusses, most of the time I do not have any middle support. The trusses span from exterior wall to exterior wall. But sometimes I will have interior walls. The wall is non load bearing wall, but in reality these non load bearing walls does take load. Loads always go to the stiffest path. So when we transition from area with interior wall to the great room, technically speaking I will have a big difference in deflection? What makes this any different?

 

[XR250]

The plywood sheathing is a roof 'member' and should be held to at least the same standard.

I get a lot of calls about this issue - where floor framing change direction. Not 3", but even 1/4" is noticeable. One reason to keep your deflections in floor framing pretty tight and not to change joist direction if at all possible.

 

[phamENG]

DoubleStud - no, I would never allow 3" deflection. That's wayyyy too much. I look at deflection along the truss, deflection of anything perpendicular (sheathing, apparent ridge, etc.), and adjacent framing of variable spans.

If standard deflections along the member won't cut it, I specify hard limits. Then if it's a change order is a fight between the GC and the truss plant.

XR- I sometimes change the framing direction to alleviate the issue. The beam needs to be matched to the joist performance as closely as possible, but then you have deflection compatibility along the transition. You might get a slight undulation in the floor, but it's better than a tear.

 

[XR250]

The beam needs to be matched to the joist performance as closely as possible, but then you have deflection compatibility along the transiti

Most of mine are walls or crawlspace girders - which basically have no deflection.

 

[JAE]

In cases where we have a stiff element (end wall of a gable roof) we try to specify a camber in the longer adjacent members so that "at rest" there is little of small/acceptable deflection.

 

[RontheRedneck]

DoubleStud said: "...are you all OK with 3” deflection on the common trusses?"

It depends on the type of building. It a post frame with no ceiling - No big deal. But in a building with drywall ceilings and walls - No way is it acceptable in my not-at-all humble opinion.

 

[XR250]

Can't imagine what is happening at the exterior wall where the subsequent horizontal deflection changes so quickly.

 

[DoubleStud]

I would be interested to know what ends up happening. I believe it is very hard to make the deflection a lot less with 60 ft span low profile scissor truss.

 

[phamENG]

Hard, but not impossible. One technique I've seen is to double the first truss off the wall. Has the effect of doubling the stiffness, which will ease the transition from theoretically infinite stiffness to slightly stiffer than a wet noodle just enough to make it palatable.